Tag
Alibaba's Qwen team released Qwen3.8-Flash-Next, an open-weight preview of the Qwen4 architecture that activates only 6B parameters out of 125B to reduce inference costs and address hardware limitations.
DeepSeek's peak-hour pricing indicates most users are in Asia, easing US concerns about Chinese AI dominance, while Alibaba's Qwen models show massive demand on Hugging Face, surpassing Meta.